A driving path planning method
A path planning and road technology, applied in neural learning methods, road network navigators, image enhancement, etc., can solve the problems of large amount of processing information, difficult implementation, expensive hardware, etc., and achieve the effect of strong performance and improved accuracy
Active Publication Date: 2022-04-15
NANJING UNIV OF INFORMATION SCI & TECH
View PDF3 Cites 0 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Improve the speed of information processing and enhance the safety of driving, solve the problems of expensive hardware, large amount of processing information, and difficult implementation in the prior art
Method used
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View moreImage
Smart Image Click on the blue labels to locate them in the text.
Smart ImageViewing Examples
Examples
Experimental program
Comparison scheme
Effect test
Embodiment Construction
[0076] Embodiments of the invention are described in detail below, examples of which are illustrated in the accompanying drawings. The embodiments described below by referring to the figures are exemplary only for explaining the present invention and should not be construed as limiting the present invention.
[0077] In most existing methods for planning driving paths using depth estimation, a decoding process that repeats simple upsampling operations is used, and in encoding, the latent properties of well-characterized objects cannot be fully exploited for monocular depth estimation. The present invention introduces the Laplacian pyramid structure into the decoder, and inputs the coding features into different video streams to decode the depth residual calculation to obtain information such as the outline size and distance of objects around the vehicle during driving, so as to plan driving path. figure 1 It is a work flow diagram of the present invention, combined with fi...
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More PUM
Login to View More Abstract
The invention discloses a driving path planning method, which uses a monocular camera to collect video stream data of the front and rear roads of the vehicle, and performs frame extraction; constructs an improved deep residual network to perform depth estimation calculation; and performs a depth map of the front and rear roads. Iterative matching and integration to form a 3D cloud image; scene semantic segmentation is performed simultaneously during the depth estimation operation; calculation of the distance between other vehicles, obstacles, signs and road lines captured by the front and rear road frame pictures and the vehicle; tracking For other vehicles in the frame image, calculate the minimum distance between the vehicle closest to the own vehicle and the own vehicle in each lane, and estimate the driving speed of the own vehicle; judge the driving angle of the vehicle according to the road line, and judge whether there is a sudden change according to the signs and road surface information. When an emergency occurs, adjust the driving route in time to allow the vehicle to drive according to the plan; perform path planning and vehicle control for the own vehicle from the position of the current frame to the position of the vehicle directly in front of it.
Description
technical field [0001] The invention relates to a driving path planning method, which belongs to the technical field of driving path planning during vehicle driving. Background technique [0002] In recent years, deep learning technology has developed rapidly and has achieved remarkable results in many application fields. Many scholars have conducted active research on monocular depth estimation using various codec architectures. Among them, the depth estimation of monocular images has always been an important issue in real scenes. key tasks. For example, the horizontal boundaries or positions of vanishing points can be efficiently estimated according to the statistics of depth information, which is very useful for quickly understanding a given scene. These clues have significant advantages in explaining the three-dimensional geometric layout. Therefore, inferring depth information has become a key technology in the field of autonomous driving. question. [0003] With the...
Claims
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More Application Information
Patent Timeline
Login to View More Patent Type & Authority Patents(China)
IPC IPC(8): G06T17/05G06T7/55G06T7/246G06T7/12G06T3/40G06V20/58G06V20/70G06N3/04G06N3/08G01C21/34
CPCG06T17/05G06T7/55G06T7/246G06T7/12G06T3/4038G06N3/08G01C21/34G06T2207/20016G06N3/045
Inventor 崔志强曹广喜单慧琳王兴涛孙佳琪张银胜
Owner NANJING UNIV OF INFORMATION SCI & TECH